There have been suggestions that the threshold for amygdala activity is lower in individuals with anxiety disorders than in healthy individuals. However, despite its immediate plausibility, there have been relatively few tests of this hypothesis. Specifically, there have been very few explorations of the performance of patients with anxiety disorders on measures known to implicate the amygdala. Although the high co-morbidity of Generalized Anxiety Disorder (GAD) and Social Anxiety Disorder (SAD) complicates the issue, the fact that the disorders doubly dissociate suggests that they are due to dysfunctional activity in separable neuro-cognitive systems. We would suggest that the hyper-responsive amygdala hypothesis is more likely to be linked to the explanation of GAD. In contrast, SAD may be due to reduced activation thresholds for units in a system that responds to social threat and which recruits lateral orbital frontal cortex. Thus, the current project will determine the performance of patients with GAD and SAD on measures in which the amygdala is known to play a role and also in measures that recruit lateral orbital frontal cortex and the system for social response reversal. In addition, two proposed neuro-imaging studies will directly assess neural responses in these two systems in both patient populations. The project aims to provide clear data that will inform future theorizing on the pathology implicated in these two disorders.